An initialization method for numerical studies of dense vesicle microchannel flows with efficient convergency

Understanding hemodynamics at the cellular scale through numerical simulations is crucial for elucidating mechanisms of cardiovascular disease and advancing microfluidic innovations in biomedical engineering. The initial placement of massive deformable red blood cells (RBCs) in microchannels not only casts great challenges to the feasibility of whole blood flows simulation at high hematocrit (Ht) but also decides computational efficiency for steady-state attainment. However, an effective initialization method for massive red blood cells ensuring rapid convergence in complex flow environments has yet to be established. This study presents an efficient method for placing cells randomly with a modified cell linked list algorithm and growth-collision model with required hematocrits in a wide range. Shrunken cells undergo controlled expansion while simultaneously translating and rotating under repulsive forces, achieving an adaptive spatial configuration that optimally occupies the available volume. The proposed method achieves dense RBC placement up to Ht=51%, while maintaining superior computational stability and efficiency across varying hematocrit conditions, with the steady-state attainment time being merely 9.5%–62.6% of that required by uniform distribution methods. It is validated that the growth-collision mechanism facilitates reliable initialization of different cell types in microchannels with arbitrary geometries. The method serves not only as an essential computational tool for cellular-scale blood flow simulations but also provides valuable reference for initializing dense-phase multiphase flow simulations.